Cooling fan control method and system and electric power equipment

Abstract

The invention discloses a cooling fan control method and system and electric power equipment. The control method comprises the steps that the output power of the electric power equipment, the measured temperature outside a power device of the electric power equipment and the current measured rotating speed of a cooling fan are collected; a state observer is configured, and the state observer uses the output power, the measured temperature and the measured rotating speed as input parameters to estimate and obtain the core temperature in a power device of the electric power equipment based on the thermodynamic model of the electric power equipment; a model prediction controller is configured, the model prediction controller carries out rolling optimization calculation based on a preset prediction model with the core temperature as an initial condition and the fan rotating speed of the future cooling fan as an optimization variable, and a target rotating speed of the cooling fan is generated; and adjusting the cooling fan according to the target rotating speed. According to the cooling fan control method, crossing from passive threshold value control to active predictive control is achieved, and the problems of lag, fluctuation and internal overheating risks in heat dissipation of high-power equipment are effectively solved.

Description

Technical Field

[0001] This invention relates to the field of heat dissipation equipment control technology, and in particular to a cooling fan control method, system and power device. Background Technology

[0002] With the rapid development of fast charging technology for electric vehicles and high-power charging pile equipment, the power density of power devices inside charging piles is constantly increasing, posing a severe challenge to their thermal management systems. As the core actuator of the thermal management system, the control strategy of the cooling fan directly affects the operational safety and stability of the equipment.

[0003] Existing cooling fan control schemes typically employ start-stop control based on a fixed temperature threshold, simple linear speed regulation, or traditional PID (proportional-integral-derivative) feedback control strategies. However, in practical applications, these existing technologies suffer from the following main technical drawbacks: First, the state perception is inaccurate, resulting in control blind spots. Existing control methods mainly rely on physical sensors installed on the outside of power devices or the surface of heat sinks in power equipment for direct readings. However, these external sensors cannot directly measure the true thermal state of the core heat-generating components inside the power devices (such as semiconductor junction temperature). Due to the resistive-capacitive effect in heat conduction, the externally measured temperature often lags behind the actual internal temperature, causing the control system to fail to accurately sense the internal thermal load of the equipment, which can easily lead to overheating damage or excessive heat dissipation of the devices.

[0004] Secondly, the control response is lag-prone and prone to fluctuations. Traditional PID control or threshold control is essentially "reactive adjustment," meaning the fan only adjusts its speed after a temperature rise or exceeding a threshold is detected. This passive response mechanism has significant control lag, making it difficult to handle sudden changes in power load and easily leading to temperature overshoot. Furthermore, to keep up with temperature changes, the fan speed is prone to frequent starts and stops or large fluctuations, which not only affects the stability of heat dissipation but also reduces the fan's lifespan. Summary of the Invention

[0005] The purpose of this invention is to provide a cooling fan control method, system, and power device that can achieve active, smooth, and precise control of the cooling fan in order to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides a cooling fan control method for an electrical power device, comprising: The system collects the output power of the power device, the external temperature of the power device, and the current speed of the cooling fan. A state observer is configured, which takes the output power, the measured temperature and the measured rotation speed as input parameters, and estimates the core temperature inside the power device of the electric power device based on the thermodynamic model of the electric power device; A model prediction controller is configured. The model prediction controller takes the core temperature as the initial condition and the future fan speed of the cooling fan as the optimization variable. Based on a preset prediction model, it performs rolling optimization calculations to generate the target speed of the cooling fan. Adjust the cooling fan according to the target rotation speed.

[0007] Preferably, the state observer is estimated using a Kalman filter algorithm.

[0008] Preferably, noise data of the environment in which the power device is located is also collected; When performing rolling optimization calculations, the model predictive controller uses the noise data as one of the initial conditions or constraints.

[0009] Preferably, when performing rolling optimization calculations, the objective function used by the model predictive controller is configured to perform weighted optimization on the following metrics: The temperature deviation between the measured temperature and the set temperature; The operating power consumption of the cooling fan; and The noise data.

[0010] Preferably, after generating the target rotational speed, the method further includes: Determine whether the power device is currently in silent mode and whether the target speed is stable. If so, a pseudo-random fluctuation signal is superimposed on the target speed to generate the final corrected speed, thereby reducing the noise of the cooling fan.

[0011] Preferably, the feedback correction method also includes the following: The measured temperature at the current moment is compared with the core temperature predicted in the previous control cycle, and the deviation value is calculated. The state estimation parameters of the state observer are corrected using the deviation value.

[0012] Preferably, the parameters of the thermodynamic model and the parameters of the prediction model are updated online based on the long-term operating data of the electric power device.

[0013] The present invention also provides an electric power device, which includes a device body and a main controller. The device body is provided with a cooling fan for convection heat dissipation, and the main controller controls the operating state of the cooling fan based on the cooling fan control method described above.

[0014] The present invention also provides a cooling fan control system, which includes: One or more processors; Memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including instructions for performing the cooling fan control method as described above.

[0015] The present invention also provides a computer-readable storage medium comprising a computer program that can be executed by a processor to perform the cooling fan control method as described above.

[0016] Compared with existing technologies, the cooling fan control method provided by the above technical solution firstly achieves accurate estimation of the unmeasurable internal temperature of power devices by fusing multi-source information through a state observer, thus solving the problems of measurement lag and inaccuracy of external sensors. Secondly, by utilizing the predictive capability of model predictive control, it realizes the transformation from passive response to active adjustment, effectively overcoming control lag, suppressing temperature overshoot, and avoiding drastic fluctuations in fan speed. While ensuring heat dissipation effect, it reduces noise and extends fan life, can quickly adapt to sudden changes in power load, and improves the robustness and reliability of the entire thermal management system. Attached Figure Description

[0017] Figure 1 This is a system framework diagram for executing the control method in an embodiment of the present invention.

[0018] Figure 2 This is a flowchart of a control method in one embodiment of the present invention.

[0019] Figure 3 This is a flowchart illustrating the overall execution process of the execution control method in this embodiment of the invention.

[0020] Figure 4 This is a comparison chart showing the control effect of the control method of the present invention on the target temperature compared with the traditional control method.

[0021] Figure 5 This is a comparison chart showing the effect of the control method of the present invention on the fan speed compared with the traditional control method. Detailed Implementation

[0022] To illustrate the technical content, structural features, objectives, and effects of the present invention in detail, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0023] This embodiment provides a method for controlling the cooling fan of power devices. This method is mainly applied to the thermal management systems of power devices such as high-power charging piles, inverters, or server power supplies. This embodiment aims to solve the control blind zone caused by the inability to directly measure the internal temperature of power devices in existing technologies, as well as the hysteresis problem of traditional feedback control.

[0024] The execution entity of this control method is usually the main controller in the power device or an independent thermal management control unit (e.g., a microprocessor based on the ARM Cortex-A53 or STM32H7 series).

[0025] Please refer to the following: Figure 1 and Figure 2 The cooling fan control method includes the following steps: S1: Data Acquisition.

[0026] Real-time acquisition of operating status data from electrical power equipment, specifically including: Output power acquisition: The current load status of the power device is read through current and voltage sensors and recorded as follows. This is a direct characteristic of the heat source.

[0027] Temperature acquisition and measurement: The current external physical temperature is obtained by using a temperature sensor (such as an NTC thermistor) mounted on the external substrate or heat sink surface of the power device (such as an IGBT or MOSFET module) in the power equipment, and denoted as . .

[0028] Speed ​​measurement: The actual current speed of the cooling fan is obtained through the fan's built-in speed feedback signal (such as the FG signal output by a Hall sensor), and recorded as _____. .

[0029] S2: Estimate internal temperature based on state observer.

[0030] Configure a state observer, which runs in the arithmetic unit of the main controller. The state observer uses the output power acquired in step S1. The measured temperature and the measured rotational speed As input parameters, based on a pre-constructed thermodynamic model of the electric power device (e.g., a thermal network model including nodes with thermal capacity and thermal resistance), the state observer reconstructs the internal thermal state of the system through mathematical calculations, thereby estimating the core temperature inside the power device, denoted as . .

[0031] This step uses algorithms to reveal the chip junction temperature, which is inaccessible to physical sensors, thus solving the problems of lag and attenuation in external measurements.

[0032] S3: Model predictive control rolling optimization.

[0033] Configure a Model Predictive Controller (MPC). This MPC uses the core temperature obtained in step S2. As the initial condition, the fan speed of the cooling fan within a finite future time domain (prediction time domain, e.g., the next 30 seconds) is used as the optimization variable.

[0034] Based on a pre-defined prediction model (which describes the dynamic relationship between speed, power, and temperature changes), rolling optimization calculations are performed while satisfying constraints (such as maximum fan speed and speed change rate limits). The goal of the calculation is to find an optimal future speed curve that optimizes system performance. The first value of this optimal curve is used as the target speed for generating the cooling fan at the current moment, denoted as . .

[0035] S4: Fan adjustment.

[0036] Based on the generated target rotational speed The MPC generates a corresponding PWM control signal to drive the fan drive circuit, adjusting the speed of the cooling fan to follow the target value.

[0037] According to the control method in this embodiment, firstly, it can accurately acquire the critical but intangible thermal state of the power device's interior (such as the chip junction temperature) without adding invasive sensors. By fusing multi-source information such as power (heat source), external temperature (thermal response), and rotational speed (heat dissipation conditions), the state observer overcomes the thermal conduction hysteresis and measurement errors caused by relying solely on external sensors, eliminates control blind spots, and ensures that control decisions are based on the core internal true temperature, thereby significantly improving the safety of equipment operation.

[0038] Secondly, it transforms traditional post-event feedback into proactive, pre-event predictive control. MPC can use predictive models to extrapolate temperature change trends over a future period and pre-plan the optimal fan speed trajectory. This means that the fan can intervene and adjust before the temperature actually rises, effectively eliminating the response lag of traditional control and greatly suppressing temperature overshoot, such as... Figure 4 This ensures that the equipment temperature is always maintained within a safe range.

[0039] Furthermore, MPC's rolling optimization mechanism allows for the consideration of control constraints (such as the rate of change of speed) when calculating the target speed. This means the system can avoid drastic fluctuations in fan speed and frequent start-stop cycles, such as... Figure 5By generating smooth target speed commands, this invention not only ensures the stability of heat dissipation but also significantly reduces fan mechanical wear and operating noise, thereby extending the service life of the cooling fan.

[0040] Finally, this method forms a closed-loop intelligent control system. When the load (output power) of the power equipment changes abruptly, it can quickly sense the disturbance through the power input, correct the state in conjunction with the state observer, and rapidly adjust the control strategy through MPC. This architecture gives the heat dissipation system extremely strong adaptability and anti-interference capability (robustness) to changes in operating conditions, maintaining optimal heat dissipation performance whether in steady-state operation or under severe load fluctuations. In another embodiment, the state observer is estimated using a Kalman filter algorithm.

[0041] The specific implementation steps are as follows: Establishing the state-space equations: Based on the thermal resistance-capacitance network of the electric power device, a discrete-time state-space model is established:

[0042]

[0043] in, This is a state vector containing the junction temperature inside the power device (i.e., the temperature to be estimated). And the temperature of each heat capacity node; The input vector contains the output power. ; The observation vector is the measured temperature. A, B, and C are system matrices; and These are process noise and measurement noise, respectively.

[0044] Prediction: Using the posterior state estimate from the previous time step and the current input parameters, predict the prior state estimate for the current time step.

[0045] Update: Calculate the Kalman gain K. Using the current measured temperature... The residual between the measured value and the predicted value, combined with the Kalman gain, is used to correct the prior state estimate, yielding the optimal posterior state estimate, i.e., the core temperature. .

[0046] Specifically, in the high-performance configuration, the state observer uses a standard Kalman filter, and the covariance matrices Q and R of process noise and measurement noise are calibrated experimentally.

[0047] In another embodiment, in practical application scenarios, such as charging stations in residential areas at night, noise from the charging stations is of particular concern. Therefore, the method in this embodiment further includes: Noise data collection: Noise data of the environment in which the power device is located is collected or acquired in real time using a microphone or sound pressure sensor integrated into the power device, or through a mapping model of fan speed and noise. This data is denoted as... .

[0048] When performing rolling optimization calculations in MPC, the noise data is used as one of the initial conditions or constraints.

[0049] When used as a constraint, a noise threshold is set. (For example, no more than 55dB in night mode). When solving for the optimal speed sequence, MPC will discard any errors that would cause the prediction noise to exceed [a certain threshold]. The speed is calculated to force the fan to operate within a quiet range unless the temperature exceeds the safe limit.

[0050] Furthermore, when performing rolling optimization calculations, the MPC employs a multi-objective cost function to evaluate each possible future speed sequence. This objective function is configured to perform weighted optimization on the following metrics: Temperature deviation: This is the square of the difference between the measured temperature and the set temperature (ideal operating temperature, such as 65℃).

[0051] Operating power consumption: Energy consumption related to the speed or control of the cooling fan.

[0052] Noise data: Noise penalty term that is non-linearly related to fan speed.

[0053] The system can dynamically adjust the weight coefficients of each optimization item based on the current time (e.g., day / night) or user mode (e.g., "rapid cooling mode" or "silent mode"). For example, at night, the noise weight will be significantly increased, causing the MPC to tend to choose a control strategy with lower speed and smoother changes, even if this results in a slight increase in temperature (but within a safe range).

[0054] In another embodiment, after generating the target rotational speed, the control method further includes a post-processing step: First, the main controller determines whether it is currently in silent mode (e.g., the system clock displays 22:00-06:00) and whether the target speed output by the MPC is stable (e.g., the speed change rate is less than a threshold in the last 5 seconds). ).

[0055] If the above conditions are met, the controller will superimpose a pseudo-random fluctuation signal with a small amplitude (e.g., ±50 RPM) and modulated frequency onto the target rotational speed.

[0056] The superimposed signal is used as the final corrected speed to drive the fan.

[0057] This slight fluctuation in fan speed will not significantly affect the heat dissipation effect, but it can effectively disperse the spectral energy of the fan's single-frequency noise, converting the sharp single-tone noise into broadband noise (similar to white noise), thereby reducing the subjective annoyance to the human ear.

[0058] In another embodiment, the cooling fan control method further includes the following feedback correction method: At each control cycle (time k), the current measured temperature is read. Simultaneously, the core temperature corresponding to the current moment, predicted from the previous control cycle (time k-1), is retrieved. Calculate the deviation between the two. .

[0059] Using the deviation value The state estimation parameters of the state observer are corrected. Typically, the deviation is multiplied by a correction gain matrix and then compensated into the current state estimation vector.

[0060] Through this closed-loop feedback, the system can eliminate steady-state errors caused by factors of the state observer itself or interference from unknown external heat sources, ensuring the accuracy of temperature control.

[0061] Furthermore, with the use of power devices, dust accumulation on the heat sink may increase thermal resistance, and wear on the fan bearings may lead to decreased efficiency. Therefore, the method in this embodiment also includes: Based on long-term operating data of the electric power equipment (e.g., historical records of power, temperature, and rotation speed over the past few weeks), the parameters of the thermodynamic model (e.g., heat and heat capacity) and the parameters of the prediction model are updated online.

[0062] Specifically, the main controller can run the Recursive Least Squares (RLS) algorithm to identify the current actual thermal parameters when the device is in standby or steady-state operation. When the identified parameters deviate from the preset parameters by more than a certain threshold, the relevant model parameters inside the main controller are updated. This ensures that throughout the entire lifecycle of the device, the main controller always performs optimized control based on a model that is closest to the actual physical state.

[0063] In summary, this invention discloses a cooling fan control method, such as... Figure 3 The specific workflow is as follows: Initialization: The system powers on and loads the required parameters.

[0064] Execute in a loop (period T = 1 second): S10: Data acquisition, i.e., reading the output power. External temperature measurement Fan speed Environmental noise .

[0065] S20: Based on state observer estimation, specifically, the Kalman filter estimates the junction temperature (i.e., core temperature) inside the IGBT based on the sampled input. (Higher than external measurements).

[0066] S30: MPC Optimization calculations are performed; specifically, MPC uses... Starting from this point, the temperature trend for the next 30 seconds is predicted. If the prediction shows that the junction temperature will rise rapidly to a safe threshold, MPC calculates a rapidly rising but smooth rotational speed curve, and the first point of this curve is taken as the target rotational speed. .

[0067] S40: Determine whether the current power device is in silent mode and the fan speed is stable. If yes, execute S50; otherwise, execute S50.

[0068] S50: Superimposes a pseudo-random signal, ultimately outputting 2480. A fluctuation command of 2520 RPM.

[0069] S60: Adjust the PWM duty cycle.

[0070] S70: Feedback correction, which means that if the actual temperature is found to be lower than the prediction during the next sampling cycle, the Kalman filter is finely adjusted through the feedback correction mechanism to prevent overcooling.

[0071] Through the above embodiments, the present invention achieves a leap from passive threshold control to active predictive control, effectively solving the problems of hysteresis, fluctuation and internal overheating risk in the heat dissipation of high-power equipment.

[0072] In another preferred embodiment of the present invention, an electric power device is also disclosed, which includes a device body and a main controller. The device body is provided with a cooling fan for convection heat dissipation, and the main controller controls the operating state of the cooling fan based on the cooling fan control method in the above embodiment.

[0073] This invention also discloses a cooling fan control system, which includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors. The programs include instructions for performing the cooling fan control method as described above. The processor may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, used to execute the relevant programs to achieve the functions required by the modules in the cooling fan control system of this application embodiment, or to execute the cooling fan control method of the method embodiment of this application.

[0074] The present invention also discloses a computer-readable storage medium comprising a computer program executable by a processor to perform the cooling fan control method described above. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center integrating one or more available media. The available medium can be read-only memory (ROM), random access memory (RAM), or magnetic media, such as floppy disks, hard disks, magnetic tapes, magnetic disks, or optical media, such as digital versatile discs (DVDs), or semiconductor media, such as solid-state disks (SSDs).

[0075] This application also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the aforementioned cooling fan control method.

[0076] The above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for controlling a cooling fan in an electrical power device, characterized in that, include: The system collects the output power of the power device, the external temperature of the power device, and the current speed of the cooling fan. A state observer is configured, which takes the output power, the measured temperature and the measured rotation speed as input parameters, and estimates the core temperature inside the power device of the electric power device based on the thermodynamic model of the electric power device; A model prediction controller is configured. The model prediction controller takes the core temperature as the initial condition and the future fan speed of the cooling fan as the optimization variable. Based on a preset prediction model, it performs rolling optimization calculations to generate the target speed of the cooling fan. Adjust the cooling fan according to the target rotation speed.

2. The cooling fan control method according to claim 1, characterized in that, The state observer is estimated using the Kalman filter algorithm.

3. The cooling fan control method according to claim 1, characterized in that, Noise data of the environment in which the power equipment is located is also collected; When performing rolling optimization calculations, the model predictive controller uses the noise data as one of the initial conditions or constraints.

4. The cooling fan control method according to claim 3, characterized in that, When performing rolling optimization calculations, the objective function used by the model predictive controller is configured to perform weighted optimization on the following indicators: The temperature deviation between the measured temperature and the set temperature; The operating power consumption of the cooling fan; and The noise data.

5. The cooling fan control method according to claim 1, characterized in that, After generating the target rotational speed, the process also includes: Determine whether the power device is currently in silent mode and whether the target speed is stable. If so, a pseudo-random fluctuation signal is superimposed on the target speed to generate the final corrected speed, thereby reducing the noise of the cooling fan.

6. The cooling fan control method according to claim 1, characterized in that, It also includes the following feedback correction methods: The measured temperature at the current moment is compared with the core temperature predicted in the previous control cycle, and the deviation value is calculated. The state estimation parameters of the state observer are corrected using the deviation value.

7. The cooling fan control method according to claim 1, characterized in that, Based on the long-term operating data of the electric power equipment, the parameters of the thermodynamic model and the parameters of the prediction model are updated online.

8. An electrical power device, characterized in that, The device includes a main body and a main controller. The main body is equipped with a cooling fan for convection cooling. The main controller controls the operating state of the cooling fan based on the cooling fan control method according to any one of claims 1 to 7.

9. A cooling fan control system, characterized in that, include: One or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including instructions for performing the cooling fan control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Includes a computer program that can be executed by a processor to perform the cooling fan control method as described in any one of claims 1 to 7.

Images